Toward Understanding Similarity of Visualization Techniques

2025-06-20

AI Summary

This study systematically characterizes similarity relationships among visualization techniques. Method: We propose a dual-path evaluation framework—“model-driven + expert-driven”: (1) a formalized visualization signature model enables structured, syntax-aware comparison of techniques; and (2) an online pairwise similarity judgment experiment involving 13 fundamental and advanced visualizations elicits expert consensus on perceptual and cognitive similarity. Contribution/Results: We establish the first dual-evaluation paradigm integrating symbolic modeling with human intuition, validated through qualitative–quantitative cross-verification. The work uncovers structured similarity patterns across diverse data types, yielding the first scalable, reproducible empirical benchmark and analytical framework for visualization similarity. This foundation advances visualization classification, retrieval, and intelligent recommendation—providing both theoretical grounding and methodological infrastructure for future research and applications.

Technology Category

• Computer Vision: Visual Reasoning & Symbolic Representations
• Data Mining & Knowledge Management: Data Visualization & Summarization
• Cognitive Modeling & Cognitive Systems: Symbolic Representations

Application Category

• Web Mining and Content Analysis: Web data visualization
• User Modeling, Personalization and Recommendation: Metrics for user behavior and evaluating success
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

The literature describes many visualization techniques for different types of data, tasks, and application contexts, and new techniques are proposed on a regular basis. Visualization surveys try to capture the immense space of techniques and structure it with meaningful categorizations. Yet, it remains difficult to understand the similarity of visualization techniques in general. We approach this open research question from two angles. First, we follow a model-driven approach that is based on defining the signature of visualization techniques and interpreting the similarity of signatures as the similarity of their associated techniques. Second, following an expert-driven approach, we asked visualization experts in a small online study for their ad-hoc intuitive assessment of the similarity of pairs visualization techniques. From both approaches, we gain insight into the similarity of a set of 13 basic and advanced visualizations for different types of data. While our results are so far preliminary and academic, they are first steps toward better understanding the similarity of visualization techniques.

Problem

• Understanding similarity of diverse visualization techniques
• Comparing model-driven and expert-driven similarity assessments
• Analyzing 13 basic and advanced visualization types

Innovation

• Model-driven approach defines visualization technique signatures
• Expert-driven approach assesses technique similarity intuitively
• Combines signatures and expert insights for similarity understanding